Encl (2) to COMDTINST 16010.8
Measurement Issues
Search and Rescue - Lives saved: Data is obtained from the Search and
Rescue Management Information System (SARMIS). SARMIS entries are
reviewed several times a year by program managers in order to validate the
data. This review includes a regression analysis to compare current data
with historical data, and analysis to identify aberrations. Current data
collection system slightly understates the number of lives lost, which
produces a percentage of lives saved that may be higher than the actual
percentage. The system is currently being updated to improve accuracy and
reliability. Confidence level in measure is moderate.
Search and Rescue - Property saved: Data is obtained from the Search and
Rescue Management Information System (SARMIS). SARMIS software contains a
number of verification checks to ensure accurate data entry. Entries are
reviewed at district offices as first step in validation. Program mangers
annually validate data. This review includes a regression analysis to
compare current data with historical data and a program review analysis to
identify and resolve aberrations. FY94-96 data is currently being
validated to remove discrepancies that may have skewed the data upwards.
Confidence level in measure is moderate.
Maritime Worker Fatalities: The baseline has been revised since the FY99
plan to incorporate new data analysis. It is established by a regression
curve obtained from several years of data. Fatality data is obtained the
Marine Safety Information System. Maritime employment estimates based on
data provided by National Marine Fisheries Service, Bureau of Labor
Statistics, and Mineral Management Service. Employment estimates have
fluctuated dramatically from year to year; substantial estimating error
exists, particularly in the fishing industry. Fatality data is obtained
from the Coast Guard Marine Safety Information System. Data is verified
by field offices, and checked during processing into the analysis
database. Program managers also check a sample of MSIS data against
incident case records. Confidence in measure is moderate to high.
Passenger Vessel Casualties: The baseline has been revised based on a new
data analysis. Hence, the goal target for FY99 has also changed from the
target in the FY99 Performance Plan. The baseline is established by a
regression curve obtained from several years of data. Performance is
measured against a selected year on the trendline. Data obtained from the
Coast Guard Marine Safety Information System (MSIS). Data is verified by
field offices. MSIS data is also checked during processing into the
analysis database. Program managers also check a sample of MSIS data
against incident case records. Confidence in measure is moderate.
Recreational Boating Fatalities: Data obtained from the Boating Accident
Report database, and is validated by program managers and state boating
officials. At the end of the calendar year, the Coast Guard compiles
statistics from the states' fatality data and sends the report to each
State for confirmation. State boating officials reconcile any data
discrepancies. The Coast Guard intends to normalize fatality data in the
future by developing a denominator of exposure so that the level of risk
in recreational boating can be compared from year to year without being
skewed by the increase or decrease in the number of boats or boat usage.
Confidence in the measure is high.
Measurement Issues 55
Previous Page
